immortal-skill

To turn personal data into a reusable AI persona, you distill multi-source inputs from chats, documents, and platforms into structured Skill bundles. This process extracts procedures, interaction styles, memories, and personality profiles to create a deployable digital persona.

Yes, you can ingest chat history from multiple platforms to create an AI skill. The system integrates over 12 data sources via intake protocols, supporting multi-platform data ingestion to comprehensively distill multi-source personal data into a unified persona bundle.

The best way to handle conflicting memories when distilling a digital persona is using evidence grading. The system tracks verbatim records, artifacts, and impressions, applying conflict handling to manage discrepancies and ensure the resulting AI skill maintains accurate personal data.

Yes, the persona distillation process supports ethical boundaries and governance. It enforces ethical constraints, manifest requirements, and versioning to manage artifacts, ensuring the generated AI skills operate within defined governance frameworks and evidence grading rules.

To generate and stamp an AI persona skill, clone the repository, set your working directory to the root, and execute Phase 0–6 using the immortal_cli.py script. This workflow processes multi-source personal data and outputs a versioned, manifest-stamped skill bundle.

Available persona templates for multi-source data distillation include seven distinct categories: self, colleague, mentor, family, partner, friend, and public-figure. These templates guide the extraction of procedures, memories, and personality profiles from multi-source personal data into reusable AI skills.